The items for my survey questionnaire that I have developed as a result of a qualitative thematic analysis, is unique. Unique in terms of, a similar questionnaire is not present in the literature. I have performed a modified Delphi technique involving experts and Cognitive interview involving young adults (potential respondents of the survey). I am confused about what kind of Construct validation method I can use for my survey? I am mentioning about construct validation as I have read that as a more acceptable kind of validation, please suggest me if anything else that I can do.
Really need help on this! Thanks for your time.
It sounds like you have taken the appropriate steps so far. The next question is whether you have the resources to do a pretest. If so, it would make sense to do a Confirmatory Factor Analysis to evaluate your proposed measure.
But if you are doing this for a doctorate, then you may not have the time or money for a pre-test. In that case you will need to do your validation work on the actual data, once you collect it.
Thanks a lot David L Morgan , I am done with my data collection. So any kind of pre-tet is not possible. Now if you can please help me with some article/links where I get to know how I can pull up the validation work on the actual data that would be really helpful.
Saswata - I'm a little confused with your study overall. Your process, although modified, does not sound like Delphi to me. What do you mean by unique? Survey-based research is not a conventional part of Delphi process - or is what you are asking related to a Delphi process plus a quesionnaire process added to it i.e. mixed methods?
Fowler's book, Survey Research Methods, has a section on validation.
In qualitative methods, you do not need any sort of statistical validity, but you need to show that the questions you asked were reasonable. Show how they were based on your literature review or other sources.
If you had not already collected your data, you might have done a pilot study, interviewing a small number of people and then fine-tuned your protocol based on feedback from the pilot study participants.
But a qualitative protocol is only a starting place. Interviews can deviate from the protocol to explore important ideas that come up in the interviews.
Saswata Ray - The major types of validity: face, construct, content, and predictive validity are important while determining the validity of a questionnaire. In face validity experts look at the items in the questionnaire and agree that the test is a valid measure of the concept which is being measured just on the face of it. In construct validity what is seen is "the degree to which a test measures what it claims, or purports, to be measuring." You could establish face validity followed by a pilot test, use a spreadsheet and use principal component analysis (PCA). Check the internal consistency of questions loading onto the same factors and revise the questionnaire based on information from your PCA.
- Badges
- Science topic
- Similar topics
- Statistics
- Survey